Is Your Browser Color Managed?

"nospam" wrote


| while you can't 'control' anything (nor was that ever a goal), you
| absolutely can ensure that an image is visually consistent across
| multiple devices.
|
No, you can't. As the article shows, even if you
use color management on your own devices, IE may
show a different image from Firefox. (That's the quote
you snipped. If you want to disagree with the author
that's one thing, but you could at least read the
article we're talking about.)

But getting back to the original point, I wasn't
questioning the value of color management locally.
I was only saying that once it comes to the Web
the idea of controlling what people see is not realistic.
The author mentions that sRGB should be used for
the Web. So what value does color management in
the browser have, unless you're viewing something
like a friend's art photographs with an embedded
color profile other than sRGB?

A real world example:

Say Eric has a family get-together next month. He
takes pictures, shooting in RAW with a good camera.
He then decides to post some online for family to see.
First, he's probably going to work with sRGB, since the
pictures are going online. Then he's probably going to
save to JPG, since the pictures are going online. Does
he need an embedded profile? Isn't sRGB default?
So why embed an sRGB profile?

Cousin Susan was wearing a very sexy red minidress
and he's got a picture of that to post. But something's
wrong. He remembers it tomato red. In the photo there
seems to be a shadow or blue tinge making it look
cranberry. So he adjusts the hue.

So.... he's got an image that's already dealing with a
limited color gamut, it's been adjusted to look the way
he remembers the scene, and he's dithered colors by
saving to JPG. If he has his own website he might want
to shrink the images to save on traffic cost. If he posts
to something like Dropbox, they might further compress
or shrink for the same reason.

Going online, the image has thus been downgraded in
several ways from the original shot. What advantage did
color management give him? It helped to ensure that he
saw on his monitor the most accurate possible colors
as captured by the camera. Whatever those are. He
thought Susan's dress needed to be altered. Was that
a problem with his eyesight? Or was the shot tinged?
Or was he so taken by Susan's behind that he imagined,
in "hindsight", let's say, that her dress was brighter than
it actually was?

No matter. The image goes online. Now 20 family members
see it. It's unlikely that even one of those people has
installed the color profile for their monitor, much less
calibrated their monitor with an external device. They
have different browsers, different eyes, different OSs,
different monitors. They're all looking at a notably
downgraded version of an altered photo of Susan's
dress. What purpose did it serve that Eric calibrated
his monitor? Almost none. It only helped him to get
the colors he wanted for his own eyes, on his own
computer, as he looked at the image of Susan colored
by his own imagination of what he saw at the party. And
as the article author pointed out, even on his machine,
his color managed browser is probably not showing him
the exact same colors that Photoshop is showing him.

Then his cousin Ed writes and says, "Nice pictures.
But how did Susan's orange dress come out red?"
Where's the discrepancy? It's anyone's guess.

You're trying to achieve an absolute objectivity where
none exists. With color management locally you can
achieve some degree of correlation, but you can't
translate that to other devices and software, and
on the Web you've already settled for a relatively
low quality image where exact color matching is not
very relevant.

"nospam" wrote

| You just snipped the quote making exactly the
| opposite point. Read the article and look at the
| comparison pictures. 3 yellow cars, all different
| hues. 3 different color-managing browsers. Why
| is this simple point so hard to grasp?
|
| looks like *you* need to reread it

Yet another informative, insightful and well-argued
point. You're a fountain of enlightenment.

In article , Mayayana
wrote:

| while you can't 'control' anything (nor was that ever a goal), you
| absolutely can ensure that an image is visually consistent across
| multiple devices.
|
No, you can't.

you clearly can't, but others sure as hell can.

As the article shows, even if you
use color management on your own devices, IE may
show a different image from Firefox. (That's the quote
you snipped. If you want to disagree with the author
that's one thing, but you could at least read the
article we're talking about.)

only slightly, and that's because ie uses a different cme.

the point of the article, which you still don't understand, is that the
yellow cars are yellow and not purple.

But getting back to the original point, I wasn't
questioning the value of color management locally.
I was only saying that once it comes to the Web
the idea of controlling what people see is not realistic.

again, the word you want is not 'control' and it's absolutely is
realistic.

The author mentions that sRGB should be used for
the Web. So what value does color management in
the browser have, unless you're viewing something
like a friend's art photographs with an embedded
color profile other than sRGB?

historically, srgb has been suggested for the web because most people
*aren't* colour managed.

however, a significant number of people have a calibrated workflow
(particularly photo pros and enthusiasts) and/or a a wide gamut display
(possibly more than one), therefore they can benefit from images that
are *not* srgb.

A real world example:

one that you made up, so it's not real world.

Say Eric has a family get-together next month. He
takes pictures, shooting in RAW with a good camera.
He then decides to post some online for family to see.
First, he's probably going to work with sRGB, since the
pictures are going online. Then he's probably going to
save to JPG, since the pictures are going online. Does
he need an embedded profile? Isn't sRGB default?
So why embed an sRGB profile?

actually, knowing what software he uses, he's likely going to work with
melissa, possibly converting for the masses, although he might export
as something better.

and i'm quite sure you won't grok what i meant. others, however, will,
particularly eric.

Cousin Susan was wearing a very sexy red minidress

yes, she certainly was.

and he's got a picture of that to post.

so does everyone else.

But something's
wrong. He remembers it tomato red. In the photo there
seems to be a shadow or blue tinge making it look
cranberry. So he adjusts the hue.

this is where you go off the rails.

So.... he's got an image that's already dealing with a
limited color gamut, it's been adjusted to look the way
he remembers the scene, and he's dithered colors by
saving to JPG. If he has his own website he might want
to shrink the images to save on traffic cost. If he posts
to something like Dropbox, they might further compress
or shrink for the same reason.

jpg doesn't necessarily mean dithered, nor is it about intentionally
downgrading images.

Going online, the image has thus been downgraded in
several ways from the original shot. What advantage did
color management give him?

accurate colours, for everyone viewing the photos.

the rest of your silly made up story snipped.

go read a book or two on colour management and try to learn something
for a change.

On May 25, 2017, Mayayana wrote
(in article ):


"nospam" wrote

while you can't 'control' anything (nor was that ever a goal), you
absolutely can ensure that an image is visually consistent across
multiple devices.
No, you can't. As the article shows, even if you
use color management on your own devices, IE may
show a different image from Firefox. (That's the quote
you snipped. If you want to disagree with the author
that's one thing, but you could at least read the
article we're talking about.)

Let’s start with the understanding that none of us, not even the most
fastidious of the color managed workflow fiends among us, can control the
viewing environment an online shared image is viewed on. It is of no
consequence if the image is viewed on a smart phone, a tablet, a Mac, or a
WIN machine, regardless of viewing software. If the creator of the image
edits with a color managed workflow, he/she should be confident that he/she
has shared an image faithful to his/her intentions.

Producing prints faithful to the edited/adjusted intent, is a totally
different proposition and shouldn’t be a part of this particular
discussion. Perhaps we can return to that as a separate subject, even though
it has been discussed in this NG many times before.

But getting back to the original point, I wasn't
questioning the value of color management locally.
I was only saying that once it comes to the Web
the idea of controlling what people see is not realistic.
The author mentions that sRGB should be used for
the Web. So what value does color management in
the browser have, unless you're viewing something
like a friend's art photographs with an embedded
color profile other than sRGB?

A real world example:

A real world example for me would entail shooting in RAW, adjusting and
editing that RAW file using LR, or ACR+PS in ProPhoto RGB, or one or another
of the apps I might be using. Once the edit/adjustment was complete to my
satisfaction I would export using the export dialog of the controlling
software. That is usually an export to JPEG with sRGB embedded.

When shared online I know for most, that the recipient’s OS + browser of
choice, and monitor used is not going to make an awkward extrapolation from
Adobe RGB, or ProPhoto RGB to sRGB. The image delivered should be consistent
with my intent. However, I can’t be responsible for an uncalibrated monitor
at the receiving end, so if the online viewer reports an inconsitancy he/she
perceives, I can make a reasonable assumption that there might well be an
issue with their viewing environment, not mine.

When it comes to me viewing an image/artwork, I have no idea of the workflow
they employ, the colorspace they used for editing/adjustments, or the
colorspace they used for the shared file. If they used something other than
sRGB there is a possibility that the results I see might have some
inconsistencies, if I’m lucky they might not. What I know for sure is the
browsers I use, Safari, and Chrome are going to deliver an sRGB image which
might choke on a file with wider gamut colorspace. It might be OK, or it
might be awful.

Say Eric has a family get-together next month. He
takes pictures, shooting in RAW with a good camera.
He then decides to post some online for family to see.
First, he's probably going to work with sRGB, since the
pictures are going online. Then he's probably going to
save to JPG, since the pictures are going online. Does
he need an embedded profile? Isn't sRGB default?
So why embed an sRGB profile?

You convert to sRGBfrom a wide gamut colorspace, and embed it because sRGB is
going to be the default for the recipient, and if A-RGB or ProPhoto RGB is
delivered there is a good possiblity all sorts of inconsistancies will be
seen. Most likely banding rather than serious color discrepancies.

Cousin Susan was wearing a very sexy red minidress
and he's got a picture of that to post. But something's
wrong. He remembers it tomato red. In the photo there
seems to be a shadow or blue tinge making it look
cranberry. So he adjusts the hue.

So.... he's got an image that's already dealing with a
limited color gamut, it's been adjusted to look the way
he remembers the scene, and he's dithered colors by
saving to JPG. If he has his own website he might want
to shrink the images to save on traffic cost. If he posts
to something like Dropbox, they might further compress
or shrink for the same reason.

That is a different issue, but for the most part Dropbox and Adobe CC seem to
deliver files they have not sat on. What is done on a personal web site or
blog is something else again.

Going online, the image has thus been downgraded in
several ways from the original shot. What advantage did
color management give him? It helped to ensure that he
saw on his monitor the most accurate possible colors
as captured by the camera. Whatever those are. He
thought Susan's dress needed to be altered. Was that
a problem with his eyesight? Or was the shot tinged?
Or was he so taken by Susan's behind that he imagined,
in "hindsight", let's say, that her dress was brighter than
it actually was?

No matter. The image goes online. Now 20 family members
see it. It's unlikely that even one of those people has
installed the color profile for their monitor, much less
calibrated their monitor with an external device. They
have different browsers, different eyes, different OSs,
different monitors. They're all looking at a notably
downgraded version of an altered photo of Susan's
dress. What purpose did it serve that Eric calibrated
his monitor? Almost none. It only helped him to get
the colors he wanted for his own eyes, on his own
computer, as he looked at the image of Susan colored
by his own imagination of what he saw at the party. And
as the article author pointed out, even on his machine,
his color managed browser is probably not showing him
the exact same colors that Photoshop is showing him.

Then his cousin Ed writes and says, "Nice pictures.
But how did Susan's orange dress come out red?"
Where's the discrepancy? It's anyone's guess.

You're trying to achieve an absolute objectivity where
none exists. With color management locally you can
achieve some degree of correlation, but you can't
translate that to other devices and software, and
on the Web you've already settled for a relatively
low quality image where exact color matching is not
very relevant.

Yup! ...and that isn’t worth getting an ulcer over.

--

Regards,
Savageduck

In article ,
Tony Cooper wrote:

I am still baffled by this type of thinking. The viewer doesn't have
any idea at all what you intended. How can the viewer report an
inconsistency of unknown values?

The only way to get the capture presented to the viewer the way you
intended it to be perceived is with a high quality print.
--
teleportation kills

On Thu, 25 May 2017 17:06:30 -0700, Savageduck
wrote:

--- snip ---

Is your computer screen calibrated?

Do you have the necessary plugins?

Maybe it's me, but anyone on Windows can test
this easily enough. You could download his
DSF4740-E.jpg and see what you think.

I would like to but I can't find the original link.

Since that is my JPEG, I am happy to oblige;
https://www.dropbox.com/s/448rl27c57zsiye/DSF4740-E.jpg

Maybe
they'll look exactly the same to you. But the
saturation and sharpness clearly look different
to me. You're not curious to know whether different
software might convey such differences? (Probably
having nothing to do with color management.)

Well! That was interesting!

I've downloaded the image via Dropbox and also directly as a JPG.

I then loaded the JPG into Photoshop and also the Windows viewer.
After a certain amount of twiddling I finished up with three images
almost exactly the same size, side by side on the screen. The
Photoshop and Firefox/Dropbox images were so similar that for
practical purposes they were identical. No doubt determined pixel
peeping would determine differences. The Windows viewer showed more
detail in the shadows and the greens were somewhat greener. I thought
it looked the best overall.

I twice tried to take a screen print and dump it into a Photoshop file
but twice I got nothing. In fact, the first time I tried it the
computer locked up and I had to resort to turning it off. Yet on other
occasions I have had no problems doing a screen dump. Most peculiar.

I notice that the colour space used by Savageduck was sRGB while the
screen on which I was using things was set to AdobeRGB. The screens
make use of an internal color matrix rather than relying on something
inside the computer. I am wondering whether or not Photoshop and
Firefox are paying attention to Savageduck's color profile while
Windows is just pouring it into my screen which displayed it as
AdobeRGB.
--

Regards,

Eric Stevens

In article ,
Tony Cooper wrote:

On Fri, 26 May 2017 06:49:01 +0200, android wrote:

In article ,
Tony Cooper wrote:

I am still baffled by this type of thinking. The viewer doesn't have
any idea at all what you intended. How can the viewer report an
inconsistency of unknown values?

The only way to get the capture presented to the viewer the way you
intended it to be perceived is with a high quality print.

This discussion has been about calibrated monitors, and the Duck's
comments were about consistency in viewing an image on different
monitors.

I'm aware of the topic of the discussion and my comment stands.
--
teleportation kills

On May 25, 2017, Tony Cooper wrote
(in ):

On Thu, 25 May 2017 21:03:45 -0700, Savageduck
wrote:

On May 25, 2017, Mayayana wrote
(in article ):


"nospam" wrote

while you can't 'control' anything (nor was that ever a goal), you
absolutely can ensure that an image is visually consistent across
multiple devices.
No, you can't. As the article shows, even if you
use color management on your own devices, IE may
show a different image from Firefox. (That's the quote
you snipped. If you want to disagree with the author
that's one thing, but you could at least read the
article we're talking about.)

Let’s start with the understanding that none of us, not even the most
fastidious of the color managed workflow fiends among us, can control the
viewing environment an online shared image is viewed on. It is of no
consequence if the image is viewed on a smart phone, a tablet, a Mac, or a
WIN machine, regardless of viewing software. If the creator of the image
edits with a color managed workflow, he/she should be confident that he/she
has shared an image faithful to his/her intentions.

Producing prints faithful to the edited/adjusted intent, is a totally
different proposition and shouldn’t be a part of this particular
discussion. Perhaps we can return to that as a separate subject, even though
it has been discussed in this NG many times before.

But getting back to the original point, I wasn't
questioning the value of color management locally.
I was only saying that once it comes to the Web
the idea of controlling what people see is not realistic.
The author mentions that sRGB should be used for
the Web. So what value does color management in
the browser have, unless you're viewing something
like a friend's art photographs with an embedded
color profile other than sRGB?

A real world example:

A real world example for me would entail shooting in RAW, adjusting and
editing that RAW file using LR, or ACR+PS in ProPhoto RGB, or one or another
of the apps I might be using. Once the edit/adjustment was complete to my
satisfaction I would export using the export dialog of the controlling
software. That is usually an export to JPEG with sRGB embedded.

When shared online I know for most, that the recipient’s OS + browser of
choice, and monitor used is not going to make an awkward extrapolation from
Adobe RGB, or ProPhoto RGB to sRGB. The image delivered should be consistent
with my intent. However, I can’t be responsible for an uncalibrated
monitor
at the receiving end, so if the online viewer reports an inconsitancy he/she
perceives, I can make a reasonable assumption that there might well be an
issue with their viewing environment, not mine.

I am still baffled by this type of thinking.

Why?

The viewer doesn't have any idea at all what you intended.

Agreed. However, if all of your work is done in a color managed environment
converted to sRGB on export and JPEG conversion for online sharing, the
viewer should see an image as you intended. For the most part they might well
see an image which is not 100% identical to the original edit, but close
enough that any subtle differences will be irrelevant.

How can the viewer report an inconsistency of unknown values?

All the viewer can do is accept that the poster of the image has made
adjustments to his/her liking for that particular subject/scene. If there is
something wrong such as a color tone, saturation level, banding in color
fields, etc., to the viewer’s eye, it is worth commenting on, and querying
the issue. That way the image creator can at least clarify their intent
and/or methodology.

There can only be consistency with a standard. Your output is the
standard in this case, and your actual output is what you see on your
monitor.

Agreed.

I can't see your monitor, so I have no idea if the image on
my monitor is consistent with that. I can't report an inconsistency
when I have no standard to use to compare.

You can certainly report that something about the image doesn’t look right
to you. That opens the discussion, and via feedback, response, and
constructive criticism an answer might be reached. Provided individuals on
either end of the exchange are open to that discussion without entering a
flame war.

All the viewer can report is whether or not he likes the rendition. If
he doesn't like it, it is not at all indicative that it's not what you
intended. He may be viewing exactly what you intended, and still not
like it.

First establish that the image as shared is as intended.
After that there is only individual taste, and we all know there is no
accounting for taste. In this NG that happens all the time, it can be
frustration when, what is a perfectly good image is ruined when shared simply
due to resizing and heavy JPEG compression, which has nothing to do with a
color managed workflow, but presents an online image which is nothing like
the original edit.

Conversely, he may think it's a brilliant rendition while looking at
something completely different from your intended appearance.

Then accept the praise, because you are ignorant of the viewer’s
perception, and you can only assume that the image was delivered as you
intended. Unless the viewer starts talking about the pink foliage, and green
sky.

....and if you were shooting IR all bets are off.

--

Regards,
Savageduck

On May 25, 2017, android wrote
(in ):

In ,
Tony Cooper wrote:

I am still baffled by this type of thinking. The viewer doesn't have
any idea at all what you intended. How can the viewer report an
inconsistency of unknown values?

The only way to get the capture presented to the viewer the way you
intended it to be perceived is with a high quality print.

....and that might be a solution, but who here is prepared to produce high
quality prints to mail around the globe for a Usenet discussion?
I might well send proofs to a client, or friends or add specific prints. I
might, and I have sent select prints to family members, or friends.

--

Regards,
Savageduck

In article ,
Tony Cooper wrote:

On Fri, 26 May 2017 07:10:48 +0200, android wrote:

In article ,
Tony Cooper wrote:

On Fri, 26 May 2017 06:49:01 +0200, android wrote:

In article ,
Tony Cooper wrote:

I am still baffled by this type of thinking. The viewer doesn't have
any idea at all what you intended. How can the viewer report an
inconsistency of unknown values?

The only way to get the capture presented to the viewer the way you
intended it to be perceived is with a high quality print.

This discussion has been about calibrated monitors, and the Duck's
comments were about consistency in viewing an image on different
monitors.

I'm aware of the topic of the discussion and my comment stands.

I didn't disagree with your comment, but it adds nothing to the
discussion.

Sure it does, and it ends it... :-))

However, there is a second way to determine consistency: two or more
people, one of which provides the standard, with their own laptops
sitting in the same room under the same conditions. Two or more
screens can be compared.

There are no two identical screens. As said:

The only way to get the capture presented to the viewer the way you
intended it to be perceived is with a high quality print.
--
teleportation kills